ymcmr-logo-new Beta

On the Comparison of Ordinary Linear and Poisson Log-linear Model for Residential Mortgage Loans

Tsung-Hao CHEN*

Assistant Professor, Department of Business Administration, Shu-Te University, Kaohsiung city, Taiwan, R.O.C.
* Corresponding author: thchen@stu.edu.tw

Keywords: Residential Mortgage Loans, Default, Tolerance, Poisson Log-linear Model

When the response variable had a normal distribution we used ordinary linear regression model to analyze data set by linking a set of explanatory variables. If the response variable is count data, counts are all positive integers, a Poisson regression model is suitable for the discrete, non-negative integer values and highly-skewed distribution of residential mortgage loans data.
In this study, we examine the data of residential mortgage loans, its shape look like Poisson distribution. So we test the distribution of scores by the Kolmogorov-smirnov one-sample test, the results indicate that the data came from a population with the Poisson distribution. Then detecting multicollinearity, we find no multicollinearity is evident by tolerance and variance inflation factor (VIF).
Although many researchers analyzed count data by ordinary linear regression, comparing finally above two previous regression analyses, all the estimate values of ordinary linear regression model are large Poisson loglinear model, but intercept is not significant and illegitimate. Therefore, we find that Poisson loglinear model has the advantage of being stably suitable for the discrete data of residential mortgage loans.

An Empirical Investigation of Mediating Effect on Corporate Social Responsibility and Financial Performance: From Intangible Assets Perspective


Associate Professor, Department of Accounting and, Centre for Environmental Accounting Research, Feng Chia University, Taiwan, R.O.C.
TEL: +886-4-24517250 ext.4231
E-Mail: shchang@fcu.edu.tw

Feng-Yi HSU

Ph.D. Program in Business, Feng Chia University, Taiwan, R.O.C.
TEL: +886-910005746
E-Mail: fengyi@mail.nmmst.gov.tw

Hsiang-Ju CHEN

Associate Professor, Department of Business Administration, De-Lin Institute of Technology, Taiwan, R.O.C.
TEL: +886-2-22733567 ext.5108
E-Mail: hjc526@yahoo.com.tw

Keywords: Corporate Social Responsibility, Tobin’s_Q, Intangible Assets, Mediating Effect

Corporate social responsibility (CSR) becomes a major issue in recent years and the correlation between CSR and corporate financial performance (CFP) is one of the research interests in related literature. However, the conclusion didn’t reach consensus. This study gets ride of traditional statistical model and utilizes intangible assets as mediating variable instead to discuss whether the mediating effect can effectively promote both CSR and CFP.
Publicly traded companies to be awarded the Corporate Social Responsibility prize of Global Views Monthly and the Excellence in Corporate Social Responsibility prize of CommonWealth for the period spanning 2007-2011 were examined as samples in this study. Apart from the difference in evaluation indicators or questionnaire survey, companies which endeavor to fulfill social responsibility would contribute to promote corporate image and to increase intangible assets that providing long-term benefits.
Relationship between CSR and CFP was examined from the perspective of intangible assets. Results show that no significant influence of CSR on CFP, but CFP indeed significant influenced CSR and therefore increase its social responsibility. Further, the relationship between CSR and CFP can be improved by the mediating effect of intangible assets.

Impacts of Knowledge Leadership and the Characteristics of Organizational Structure on Employee Learning Motivation in the Cultural and Creative Industries

Yuan-Cheng CHANG

Ph D Candidate, Graduate Institute of Educational Entrepreneurship and Management, National University of Tainan (Taiwan)
E-Mail: chen11777@hotmail.com


Ph D, Assistant Professor, Department of Recreation and Sports Management, Shu-Te University (Taiwan)
Address: No.59, Hengshan Rd., Yanchao Dist., Kaohsiung City, 82445 Taiwan, R.O.C.
TEL: +886-7-615-8000 ext.6412
Corresponding author: christinechiu@stu.edu.tw

Yuan-Duen LEE

Ph D, Professor, Graduate School of Business and Operations Management, Chang Jung Christian University, Taiwan.
E-Mail: ydlee@mail.cjcu.edu.tw

Keywords: Knowledge Leadership, Characteristics of Organizational Structure, Learning Motivation

This study explores the influence of knowledge leadership and the characteristics of organizational structure on employee learning motivation in the cultural and creative industries in Taiwan. A questionnaire survey of related enterprises was conducted with stratified sampling, resulting in a collection of valid samples from 36 firms, with 76 copies completed by mid- and high-level managers and 309 by other employees. A hierarchical linear modeling (HLM) analysis of the data was carried out, and the results indicate that in terms of the characteristics of organizational structure, neither formalization nor centralization has a significant influence on employee learning motivation. The practice of knowledge management as an aspect of knowledge leadership has a significant, negative influence on employee learning motivation. In contrast, creating a learning environment has a significantly positive influence on employee learning motivation. In terms of knowledge leadership, knowledge sharing and evaluation have a negative moderating effect on the relationship between a centralized organization and employee learning motivation. Creating a learning environment has a positive moderating effect on the relationship between a formalized organization and employee learning motivation. The promotion of knowledge management has a negative moderating effect on the relationship between a centralized organization and employee learning motivation.

Does Noise Matter on Hedging Effectiveness? Evidence from Taiwan

Juping WU

Assistant Professor of Department of Finance, Shu-Te University, Taiwan, R.O.C.

Ren-Jean LIOU

Associate Professor and Chairman of Department of Computers and Communications, National Pingtung Institute of Commerce, Taiwan
TEL: +886-7-6158000-3212
E-Mail: juping@stu.edu.tw

Keywords: Hedging Strategy; Noise; Optimal Hedge Ratio; Emerging Market; Information Asymmetry

This paper explores the relationship between different price components and optimal hedge ratios. The conventional estimation methods obtain optimal hedge ratios using original price series. In this paper, I propose associating original, pure-informed, and noisy added signals to estimate hedge ratios. Hedge effectiveness is obtained by incorporating most recommended methods, such as OLS, EC, and GARCH model. This paper proposes a wavelet approach to decompose the price series into informed and noise components and evaluate the hedging performance with the hedge ratios computed based on the two components. Using the Taiwan stock index and futures with three alternative models, the empirical results support the merit of this decomposition for the spot-futures hedge. The results show that, in a daily hedging strategy (i.e., short-term hedging), optimal hedge ratios are estimated using noise-added information. A de-noised signal better estimates long-term hedging effectiveness than does that of original information. Both low and high frequency portions of the signal are important for hedging purposes depending horizon. Noise is often in the form of market microstructure nature and serially correlated but no treatment is provided in the past researches. The practical implications for practitioners are considering that noise does affect hedging performance, particularly in the short run.

The Application of Discriminants to Avoid Erroneous GM(1,1) Prediction


Professor of Department of Industrial Management, I-Shou University, No.1, Sec. 1, Syuecheng Rd., Dashu District, Kaohsiung City 84001,Taiwan, R.O.C.
TEL: +886-7-6577711 ext.5522; FAX: +886-7-6578536
E-Mail: EddyChen@isu.edu.tw
To whom all correspondence should be addressed

Shou-Jen HUANG

PhD graduate of Department of Industrial Management, I-Shou University

Keywords: Grey Prediction, Singular Phenomena, Discriminants, Currency Exchange

The intrinsic defect of GM(1,1) that grey development coefficient a equals to zero will cause a significant and meaningless prediction errors in subsequent calculations. The researchers should be cautious when apply GM(1,1). Before applying GM(1,1), a pretest of raw data should be performed in order to avoid erroneous forecasting. In this research, a case of currency monthly exchange rate between HK dollar and US dollar from August, 2011 to January, 2012 has been shown to emphasize the importance of pretest of raw data.